INDEX

Explanations

terms related to operational activities or structures

oai_token-act-pair · gpt-3.5-turbo

terms related to operational status and functions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

Pac

-0.75

hner

-0.74

eve

-0.71

kick

-0.71

astern

-0.69

 Pupp

-0.69

xx

-0.69

Fey

-0.68

gets

-0.68

 Neighbor

-0.67

POSITIVE LOGITS

 operational

1.22

 Oper

0.93

izational

0.88

espie

0.85

 operation

0.84

 personnel

0.83

ilitary

0.82

ational

0.81

 capability

0.80

 operative

0.78

Activations Density 0.007%

terms related to operational activities or structures

terms related to operational status and functions

No Comments

No Known Activations