INDEX

Explanations

idiomatic expressions and phrases from various contexts

oai_token-act-pair · gpt-3.5-turbo

references to outcomes that involve winning or losing in various contexts

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

-0.69

 newsp

-0.60

 repertoire

-0.58

 orche

-0.57

 referen

-0.57

cir

-0.53

vernment

-0.52

 RELE

-0.52

 Tradable

-0.52

 phased

-0.51

POSITIVE LOGITS

;}

0.97

.--

0.90

.—

0.88

0.84

.;

0.81

():

0.79

'';

0.78

.?

0.76

.#

0.76

Activations Density 1.010%

idiomatic expressions and phrases from various contexts

references to outcomes that involve winning or losing in various contexts

No Comments

No Known Activations