INDEX

Explanations

instances of the word "Doc"

oai_token-act-pair · gpt-3.5-turbo

mentions of the term "Doc" in various contexts

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

theless

-0.71

WAYS

-0.67

places

-0.66

 indignation

-0.64

 susceptibility

-0.62

committee

-0.62

mare

-0.62

 hazards

-0.61

 perme

-0.60

POL

-0.60

POSITIVE LOGITS

uments

1.14

herty

1.05

sis

1.03

Doc

1.02

ents

0.99

umen

0.94

doc

0.91

ually

0.91

ilt

0.86

ual

0.86

Activations Density 0.020%

instances of the word "Doc"

mentions of the term "Doc" in various contexts

No Comments

No Known Activations