INDEX

Explanations

phrases containing the word "mini" with negative connotations

oai_token-act-pair · gpt-3.5-turbo

references to "mini" or "micro" versions of things

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 Zimmer

-0.73

 GOODMAN

-0.71

 Petty

-0.70

 Ernst

-0.67

 Cohn

-0.66

 outnumbered

-0.66

 Polk

-0.66

 Reilly

-0.63

 Center

-0.63

 Strauss

-0.62

POSITIVE LOGITS

series

1.16

dial

1.14

repl

1.08

deal

1.06

definition

1.05

instance

1.04

analysis

1.03

batch

1.03

government

1.03

data

1.02

Activations Density 0.066%

phrases containing the word "mini" with negative connotations

references to "mini" or "micro" versions of things

No Comments

No Known Activations