INDEX

Explanations

phrases related to challenging situations or actions

oai_token-act-pair · gpt-3.5-turbo

specific single-letter prefixes or abbreviations

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

hyde

-0.83

 substitutes

-0.69

 constructs

-0.68

 Fargo

-0.68

 dwarves

-0.66

 sacrific

-0.65

 rescued

-0.64

FORE

-0.64

 promot

-0.63

 learners

-0.62

POSITIVE LOGITS

anky

1.03

agging

0.99

ithering

0.98

umbling

0.96

ashing

0.94

attering

0.94

agg

0.94

angu

0.94

acious

0.91

erb

0.91

Activations Density 0.238%

phrases related to challenging situations or actions

specific single-letter prefixes or abbreviations

No Comments

No Known Activations