INDEX

Explanations

triggers related to fragmented objects or environments

oai_token-act-pair · gpt-3.5-turbo

the preposition "of" indicating relationships or descriptions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

 Kills

-0.70

Ô

-0.70

Pause

-0.66

eus

-0.64

 Cosponsors

-0.63

iest

-0.61

beam

-0.61

 istg

-0.61

ifier

-0.60

mie

-0.60

POSITIVE LOGITS

 varying

0.82

 differing

0.76

 assorted

0.73

 dubious

0.67

tnc

0.67

sembly

0.67

 themselves

0.66

 disparate

0.66

 sorts

0.63

 prey

0.63

Activations Density 0.250%

triggers related to fragmented objects or environments

the preposition "of" indicating relationships or descriptions

No Comments

No Known Activations