INDEX

Explanations

words related to suitability or compatibility

oai_token-act-pair · gpt-3.5-turbo

terms related to suitability and appropriateness

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

 helicop

-0.76

 Pledge

-0.62

odium

-0.61

cler

-0.61

cens

-0.60

olan

-0.60

 taped

-0.60

Kah

-0.60

opus

-0.59

plane

-0.58

POSITIVE LOGITS

 suited

1.00

 candidates

0.79

 suitable

0.75

 conducive

0.71

for

0.71

fit

0.69

 elig

0.68

folios

0.67

IENT

0.67

 embodiments

0.67

Activations Density 0.076%

words related to suitability or compatibility

terms related to suitability and appropriateness

No Comments

No Known Activations