INDEX

Explanations

explicit statements or propositions

oai_token-act-pair · gpt-3.5-turbo

references to various types of statements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

elsius

-0.78

rys

-0.76

rowd

-0.72

MpServer

-0.70

rowing

-0.69

itals

-0.68

bid

-0.67

Friend

-0.66

versely

-0.66

axy

-0.66

POSITIVE LOGITS

 statements

0.94

 statement

0.90

ariat

0.88

 pronoun

0.81

 uttered

0.79

gow

0.79

ARB

0.77

Statement

0.76

 regarding

0.76

 Statements

0.74

Activations Density 0.027%

explicit statements or propositions

references to various types of statements

No Comments

No Known Activations