INDEX

Explanations

punctuation marks, specifically commas

oai_token-act-pair · gpt-3.5-turbo

interrogative phrases and conjunctions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

RM

-0.64

SourceFile

-0.61

front

-0.60

SPA

-0.59

BS

-0.58

oir

-0.58

Nap

-0.58

letters

-0.58

LF

-0.57

Actor

-0.57

POSITIVE LOGITS

into

0.65

limits

0.64

 curfew

0.63

ceptor

0.61

ttle

0.60

 tacit

0.59

 diminishing

0.56

cipl

0.56

azard

0.55

Îº

0.54

Activations Density 0.207%

punctuation marks, specifically commas

interrogative phrases and conjunctions

No Comments

No Known Activations