INDEX

Explanations

specific and precise locations or details

oai_token-act-pair · gpt-3.5-turbo

references to precision or specific details

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

rift

-0.77

 Pigs

-0.75

ker

-0.69

stown

-0.68

ipolar

-0.67

NetMessage

-0.66

 ashore

-0.66

OPLE

-0.65

ulton

-0.65

acca

-0.65

POSITIVE LOGITS

itude

0.94

 wording

0.88

 exact

0.87

ing

0.85

aneous

0.84

ions

0.83

 embodiment

0.78

 same

0.76

 opposite

0.74

itudes

0.73

Activations Density 0.014%

specific and precise locations or details

references to precision or specific details

No Comments

No Known Activations