INDEX

Explanations

phrases indicating an impression or perception

oai_token-act-pair · gpt-3.5-turbo

instances of the word "impression" and its various contexts

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

 annex

-0.68

annis

-0.68

foreseen

-0.67

fighting

-0.65

cross

-0.65

 challeng

-0.64

yne

-0.64

vez

-0.63

ccording

-0.63

cale

-0.62

POSITIVE LOGITS

 impression

1.12

 impressions

1.00

uren

0.84

eful

0.81

eless

0.76

Poc

0.72

 perceptions

0.71

 scars

0.69

uated

0.68

able

0.68

Activations Density 0.017%

phrases indicating an impression or perception

instances of the word "impression" and its various contexts

No Comments

No Known Activations