INDEX

Explanations

phrases related to taking something to a higher level or extreme

oai_token-act-pair · gpt-3.5-turbo

phrases expressing escalation or advancement to a new level

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

uates

-0.80

icipated

-0.67

/-

-0.65

 elig

-0.63

hered

-0.61

uated

-0.59

etta

-0.59

accompanied

-0.58

 withheld

-0.58

hart

-0.58

POSITIVE LOGITS

 extremes

0.91

 fruition

0.74

 stride

0.73

 lengths

0.69

 task

0.69

psy

0.68

 shred

0.64

 heart

0.63

 extrem

0.62

 pasture

0.62

Activations Density 0.069%

phrases related to taking something to a higher level or extreme

phrases expressing escalation or advancement to a new level

No Comments

No Known Activations