INDEX

Explanations

words related to creating obstacles or hindrances

oai_token-act-pair · gpt-3.5-turbo

terms related to obstruction and structure

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 thirds

-0.77

 Lans

-0.61

butt

-0.61

bes

-0.59

Ole

-0.58

fare

-0.57

 Giul

-0.56

 Teresa

-0.56

 Haram

-0.56

 ashore

-0.55

POSITIVE LOGITS

ions

2.20

ively

2.20

ive

2.14

ives

2.00

iveness

1.80

ivity

1.79

ible

1.78

ional

1.75

ivist

1.72

ors

1.71

Activations Density 0.195%

words related to creating obstacles or hindrances

terms related to obstruction and structure

No Comments

No Known Activations