INDEX

Explanations

words related to individuals or places with the prefix "Or-" followed by specific characters

oai_token-act-pair · gpt-3.5-turbo

occurrences of the substring "ang" and related variations in words

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

Reviewer

-0.68

puter

-0.61

lif

-0.60

 ISBN

-0.59

restling

-0.57

inki

-0.54

unb

-0.54

 competitor

-0.54

KR

-0.54

 vampires

-0.52

POSITIVE LOGITS

odox

0.83

 Pradesh

0.77

utan

0.74

hemy

0.74

oi

0.74

iculture

0.73

ocene

0.73

neau

0.72

emort

0.71

ĺħ

0.68

Activations Density 0.098%

words related to individuals or places with the prefix "Or-" followed by specific characters

occurrences of the substring "ang" and related variations in words

No Comments

No Known Activations