INDEX

Explanations

pronouns that refer to a specific concept being mentioned

oai_token-act-pair · gpt-3.5-turbo

instances and references of the word "it."

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

ãĤ·ãĥ£

-0.71

-0.62

Topics

-0.61

 Aluminum

-0.58

 Mans

-0.57

pi

-0.56

course

-0.56

 Carbon

-0.56

 Hanson

-0.55

Dam

-0.54

POSITIVE LOGITS

chy

0.93

 enthusi

0.88

 firsthand

0.85

 recre

0.84

 commercially

0.83

self

0.82

 recip

0.79

selves

0.79

 legitimately

0.76

 anecd

0.76

Activations Density 0.175%

pronouns that refer to a specific concept being mentioned

instances and references of the word "it."

No Comments

No Known Activations