INDEX

Explanations

words related to physical damage, wounds, or graphic descriptions of bodily harm

oai_token-act-pair · gpt-3.5-turbo

themes related to violence and its consequences

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

izont

-0.66

otiation

-0.64

elfth

-0.62

hander

-0.61

imeo

-0.59

assad

-0.59

ollower

-0.59

 Participant

-0.58

akening

-0.58

rolet

-0.57

POSITIVE LOGITS

 goodies

0.94

 fumes

0.84

 tentacles

0.84

 junk

0.84

 feces

0.81

 crap

0.80

 smells

0.79

 acron

0.79

 needles

0.79

 horrors

0.78

Activations Density 0.442%

words related to physical damage, wounds, or graphic descriptions of bodily harm

themes related to violence and its consequences

No Comments

No Known Activations