INDEX

Explanations

titles or occupations related to spokespersons or representatives

oai_token-act-pair · gpt-3.5-turbo

occurrences of the article "a" before nouns

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

eye

-0.79

frames

-0.75

EXP

-0.75

sticks

-0.74

views

-0.73

EMA

-0.72

Episode

-0.68

boats

-0.68

orders

-0.68

fn

-0.67

POSITIVE LOGITS

 longtime

1.09

 former

1.08

 spokesman

1.07

 spokesperson

1.05

 member

1.04

 descendant

0.99

 spokeswoman

0.98

 businessman

0.97

 retired

0.96

 lecturer

0.96

Activations Density 0.127%

titles or occupations related to spokespersons or representatives

occurrences of the article "a" before nouns

No Comments

No Known Activations