INDEX

Explanations

words related to preferences and choices

oai_token-act-pair · gpt-3.5-turbo

terms related to preferences

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

meal

-0.74

mans

-0.73

manship

-0.69

mberg

-0.67

Interstitial

-0.67

FORMATION

-0.67

together

-0.67

rooms

-0.66

semble

-0.64

angel

-0.64

POSITIVE LOGITS

 preferences

1.02

 favoring

0.90

 favoured

0.84

 preference

0.82

eering

0.77

ministic

0.76

 favors

0.76

fav

0.76

 Preferences

0.75

ately

0.75

Activations Density 0.039%

words related to preferences and choices

terms related to preferences

No Comments

No Known Activations