INDEX

Explanations

adjectives related to characteristics of behavior or actions

oai_token-act-pair · gpt-3.5-turbo

terms related to fairness, impartiality, and complicity

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

RAM

-0.73

bum

-0.72

 Jensen

-0.67

\\\\\\\\\\\\\\\\

-0.67

trap

-0.66

clamation

-0.65

 Grow

-0.65

hib

-0.64

reat

-0.63

 Bray

-0.62

POSITIVE LOGITS

 impartial

0.88

seys

0.88

ity

0.86

 complicity

0.79

 Cosponsors

0.77

 complicit

0.76

itous

0.74

 spectator

0.72

 postage

0.69

ences

0.69

Activations Density 0.023%

adjectives related to characteristics of behavior or actions

terms related to fairness, impartiality, and complicity

No Comments

No Known Activations