INDEX

Explanations

mentions of items or concepts being fully completed or operational

oai_token-act-pair · gpt-3.5-turbo

expressions indicating completeness or totality

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

liest

-0.89

lies

-0.77

itism

-0.68

 Parables

-0.68

mere

-0.67

 Mostly

-0.67

 Passage

-0.66

ares

-0.65

lier

-0.64

gee

-0.63

POSITIVE LOGITS

fledged

1.12

 fled

1.09

 clothed

0.96

 stocked

0.93

 realized

0.92

 realised

0.89

 formed

0.87

 immersed

0.87

 Loaded

0.86

 utilized

0.86

Activations Density 0.030%

mentions of items or concepts being fully completed or operational

expressions indicating completeness or totality

No Comments

No Known Activations