INDEX

Explanations

words related to darkness or negative forces

oai_token-act-pair · gpt-3.5-turbo

references to darkness or themes of darkness

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

ctions

-0.92

ctive

-0.82

TAIN

-0.81

kson

-0.80

ction

-0.79

lli

-0.77

IGN

-0.76

KEY

-0.72

Union

-0.71

GN

-0.71

POSITIVE LOGITS

 darkness

0.90

 shroud

0.86

 abyss

0.84

dark

0.78

iness

0.78

 reckoning

0.78

 shone

0.77

 eleph

0.76

mares

0.76

 violet

0.75

Activations Density 0.011%

words related to darkness or negative forces

references to darkness or themes of darkness

No Comments

No Known Activations