INDEX

Explanations

words related to provocation or triggering actions

oai_token-act-pair · gpt-3.5-turbo

words related to evidence or proof

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

morph

-0.79

 Halls

-0.76

Lay

-0.72

Reloaded

-0.68

 cryst

-0.68

Ago

-0.66

urgy

-0.63

ecycle

-0.62

 proficiency

-0.61

 Lama

-0.61

POSITIVE LOGITS

 prov

0.90

atform

0.87

fare

0.87

iso

0.84

irus

0.84

idential

0.76

inelli

0.76

ably

0.74

rop

0.74

itely

0.73

Activations Density 0.049%

words related to provocation or triggering actions

words related to evidence or proof

No Comments

No Known Activations