INDEX

Explanations

specific references to objects or items that are being firmly held or tightly grasped, particularly items that can be categorized as a "clutch"

oai_token-act-pair · gpt-3.5-turbo

words related to the concept of clutter or things being clustered together

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ordan

-0.84

ozy

-0.73

ItemTracker

-0.72

UFC

-0.72

DIT

-0.71

ĨĴ

-0.68

Ô

-0.66

IRO

-0.66

 attm

-0.65

 Nation

-0.64

POSITIVE LOGITS

tered

1.22

ches

1.00

tering

0.95

rial

0.88

rette

0.86

ters

0.84

ched

0.84

bread

0.82

cher

0.79

rily

0.79

Activations Density 0.049%

specific references to objects or items that are being firmly held or tightly grasped, particularly items that can be categorized as a "clutch"

words related to the concept of clutter or things being clustered together

No Comments

No Known Activations