INDEX

Explanations

names of individuals who are leading or heading a group or an activity

oai_token-act-pair · gpt-3.5-turbo

instances of the word "led," indicating leadership or direction in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Effective

-0.66

 redundancy

-0.61

FTA

-0.61

ypes

-0.59

iddler

-0.59

 competitiveness

-0.58

 srfAttach

-0.57

weather

-0.57

 DISTRICT

-0.56

 comfort

-0.55

POSITIVE LOGITS

gers

1.18

ger

1.01

ges

0.99

gling

0.98

better

0.96

ging

0.89

ership

0.82

erer

0.81

ged

0.80

gew

0.79

Activations Density 0.050%

names of individuals who are leading or heading a group or an activity

instances of the word "led," indicating leadership or direction in various contexts

No Comments

No Known Activations