INDEX

Explanations

words related to goals, assessments, or objective statements

oai_token-act-pair · gpt-3.5-turbo

terms related to objectives and goals

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

oles

-0.94

oly

-0.82

ulet

-0.80

artifacts

-0.79

 Tycoon

-0.75

olini

-0.74

buck

-0.74

alter

-0.73

ule

-0.73

ole

-0.72

POSITIVE LOGITS

ives

0.90

 Objective

0.88

 objective

0.81

IVES

0.79

 observer

0.77

ignty

0.77

imity

0.74

JECT

0.72

 objectives

0.71

 guiActiveUn

0.69

Activations Density 0.012%

words related to goals, assessments, or objective statements

terms related to objectives and goals

No Comments

No Known Activations