INDEX

Explanations

terms related to video footage

oai_token-act-pair · gpt-3.5-turbo

references to video footage

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

inence

-0.79

Ã±a

-0.74

under

-0.73

isner

-0.71

 boycot

-0.67

ggles

-0.67

neg

-0.67

 depress

-0.66

patient

-0.66

 dement

-0.64

POSITIVE LOGITS

 footage

1.40

 CCTV

1.06

 clips

1.05

 clip

0.94

 tape

0.93

 tapes

0.93

 filmed

0.88

 recordings

0.84

 taped

0.84

 cameras

0.84

Activations Density 0.015%

terms related to video footage

references to video footage

No Comments

No Known Activations