INDEX

Explanations

terms related to locations or events

oai_token-act-pair · gpt-3.5-turbo

terms related to events, positions, and significant identifiers in a narrative context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

 Sins

-0.80

 Dough

-0.75

nces

-0.74

 Languages

-0.72

 Cookies

-0.70

DF

-0.70

 Technique

-0.70

 Stain

-0.70

 Leaders

-0.69

 Guests

-0.69

POSITIVE LOGITS

1.05

-'

0.82

 fide

0.75

-[

0.73

-,

0.72

-)

0.70

âĢĳ

0.65

 lane

0.65

-$

0.65

 fund

0.64

Activations Density 0.388%

terms related to locations or events

terms related to events, positions, and significant identifiers in a narrative context

No Comments

No Known Activations