INDEX

Explanations

phrases containing the words "one-stop" or describing actions done once

oai_token-act-pair · gpt-3.5-turbo

terms related to time and duration

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

natureconservancy

-0.77

successfully

-0.72

prof

-0.71

helm

-0.71

raught

-0.69

ãĤ©

-0.69

yn

-0.68

ulkan

-0.67

mos

-0.66

enegger

-0.66

POSITIVE LOGITS

 Wonders

0.84

 increments

0.78

 anniversary

0.74

ousand

0.72

undred

0.66

 Anniversary

0.65

 shop

0.64

 Bleach

0.62

 cameo

0.60

Ans

0.60

Activations Density 0.072%

phrases containing the words "one-stop" or describing actions done once

terms related to time and duration

No Comments

No Known Activations