INDEX

Explanations

mentions or implications of concepts or ideas without explicit statements

oai_token-act-pair · gpt-3.5-turbo

terms related to tacit and implicit agreements or understandings

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ĪĴ

-0.91

ishers

-0.77

gard

-0.74

arios

-0.72

etsk

-0.72

rir

-0.71

rar

-0.71

rament

-0.70

hemy

-0.70

frey

-0.69

POSITIVE LOGITS

 tacit

1.04

 assumption

0.94

 endorse

0.84

 implicitly

0.83

icit

0.82

 complicity

0.80

 guiActiveUn

0.78

 invitation

0.78

 acknowledgment

0.78

 consent

0.78

Activations Density 0.026%

mentions or implications of concepts or ideas without explicit statements

terms related to tacit and implicit agreements or understandings

No Comments

No Known Activations