INDEX

Explanations

people or entities that are "familiar with" certain situations or topics

oai_token-act-pair · gpt-3.5-turbo

terms related to people and their societal or health conditions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

Quantity

-0.66

phrine

-0.65

PokÃ©

-0.63

OLOG

-0.60

 vortex

-0.58

 Depot

-0.57

theless

-0.57

phasis

-0.57

........

-0.56

 arrang

-0.56

POSITIVE LOGITS

iaries

0.83

ocument

0.72

viol

0.68

 unre

0.64

 themselves

0.64

 afflicted

0.61

chool

0.60

 inhabit

0.60

 deceased

0.60

inside

0.60

Activations Density 0.446%

people or entities that are "familiar with" certain situations or topics

terms related to people and their societal or health conditions

No Comments

No Known Activations