INDEX

Explanations

mentions of self-related concepts or actions

oai_token-act-pair · gpt-3.5-turbo

terms related to self-identity or self-referential concepts

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

 Nights

-0.79

 XIII

-0.77

 Ashe

-0.75

 Rouge

-0.72

ī

-0.71

SHIP

-0.71

 Syndicate

-0.70

IUM

-0.68

 GOODMAN

-0.64

LAT

-0.64

POSITIVE LOGITS

destruct

1.30

same

1.06

ridges

0.99

 destruct

0.97

lessly

0.96

conscious

0.96

 explanatory

0.94

contained

0.91

 esteem

0.86

upload

0.84

Activations Density 0.034%

mentions of self-related concepts or actions

terms related to self-identity or self-referential concepts

No Comments

No Known Activations