INDEX

Explanations

names, specifically those starting with "Den"

oai_token-act-pair · gpt-3.5-turbo

occurrences of the term "Den" 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 @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

natureconservancy

-0.71

ties

-0.69

 behav

-0.68

mentation

-0.67

ulatory

-0.65

mph

-0.64

EED

-0.62

 Typh

-0.62

olicy

-0.62

rious

-0.61

POSITIVE LOGITS

omination

1.11

omin

0.98

izen

0.92

zel

0.90

izens

0.88

agram

0.88

arius

0.84

elson

0.83

holm

0.83

ega

0.81

Activations Density 0.036%

names, specifically those starting with "Den"

occurrences of the term "Den" in various contexts

No Comments

No Known Activations