INDEX

Explanations

words related to strong connections or inevitable outcomes

oai_token-act-pair · gpt-3.5-turbo

instances of the word "bound."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

issance

-0.89

sych

-0.77

soDeliveryDate

-0.71

ciation

-0.70

 fielded

-0.66

HAEL

-0.65

STEM

-0.65

ETHOD

-0.64

amera

-0.64

oples

-0.63

POSITIVE LOGITS

less

0.98

fold

0.96

 bound

0.96

lessly

0.95

unin

0.93

binding

0.90

bound

0.88

ging

0.86

sym

0.86

gling

0.82

Activations Density 0.021%

words related to strong connections or inevitable outcomes

instances of the word "bound."

No Comments

No Known Activations