INDEX

Explanations

references to comparing or connecting things, such as "watching X to understand Y" or "this movie reminded me of another film."

oai_token-act-pair · gpt-3.5-turbo

references to movies and their impact on viewers

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 Species

-0.68

ymm

-0.68

DOE

-0.65

ebus

-0.65

 turf

-0.65

osi

-0.64

 Cutter

-0.63

 Gibbs

-0.63

 USPS

-0.63

 metals

-0.62

POSITIVE LOGITS

 aloud

1.02

 firsthand

0.90

 binge

0.88

 comprehension

0.86

 headphones

0.82

 obsess

0.77

resso

0.77

ousands

0.76

watching

0.75

 nostalg

0.75

Activations Density 0.652%

references to comparing or connecting things, such as "watching X to understand Y" or "this movie reminded me of another film."

references to movies and their impact on viewers

No Comments

No Known Activations