INDEX

Explanations

comparative expressions about actions or decisions

oai_token-act-pair · gpt-3.5-turbo

discussions about preferences and opinions regarding choices or options

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

 ÂŃ

-0.98

ewitness

-0.76

Washington

-0.76

Federal

-0.75

 telephone

-0.74

âĢ¦âĢ¦âĢ¦âĢ¦âĢ¦âĢ¦âĢ¦âĢ¦

-0.71

 Washington

-0.71

âĢİ

-0.70

——

-0.69

vernment

-0.69

POSITIVE LOGITS

 gameplay

1.22

 combos

1.15

 playthrough

1.15

 devs

1.13

 playable

1.09

ner

1.06

 mech

1.04

 setups

1.01

 alot

1.01

 overpowered

1.01

Activations Density 1.121%

comparative expressions about actions or decisions

discussions about preferences and opinions regarding choices or options

No Comments

No Known Activations